{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,6,18]],"date-time":"2026-06-18T02:38:19Z","timestamp":1781750299381,"version":"3.54.5"},"reference-count":64,"publisher":"MIT Press","license":[{"start":{"date-parts":[[2022,11,8]],"date-time":"2022-11-08T00:00:00Z","timestamp":1667865600000},"content-version":"vor","delay-in-days":311,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":["direct.mit.edu"],"crossmark-restriction":true},"short-container-title":[],"published-print":{"date-parts":[[2022,11,7]]},"abstract":"<jats:title>Abstract<\/jats:title><jats:p>Generalizing dialogue state tracking (DST) to new data is especially challenging due to the strong reliance on abundant and fine-grained supervision during training. Sample sparsity, distributional shift, and the occurrence of new concepts and topics frequently lead to severe performance degradation during inference. In this paper we propose a training strategy to build extractive DST models without the need for fine-grained manual span labels. Two novel input-level dropout methods mitigate the negative impact of sample sparsity. We propose a new model architecture with a unified encoder that supports value as well as slot independence by leveraging the attention mechanism. We combine the strengths of triple copy strategy DST and value matching to benefit from complementary predictions without violating the principle of ontology independence. Our experiments demonstrate that an extractive DST model can be trained without manual span labels. Our architecture and training strategies improve robustness towards sample sparsity, new concepts, and topics, leading to state-of-the-art performance on a range of benchmarks. We further highlight our model\u2019s ability to effectively learn from non-dialogue data.<\/jats:p>","DOI":"10.1162\/tacl_a_00513","type":"journal-article","created":{"date-parts":[[2022,11,8]],"date-time":"2022-11-08T21:00:01Z","timestamp":1667941201000},"page":"1175-1192","update-policy":"https:\/\/doi.org\/10.1162\/mitpressjournals.corrections.policy","source":"Crossref","is-referenced-by-count":8,"title":["Robust Dialogue State Tracking with Weak Supervision and Sparse Data"],"prefix":"10.1162","volume":"10","author":[{"given":"Michael","family":"Heck","sequence":"first","affiliation":[{"name":"Heinrich Heine University D\u00fcsseldorf, Germany. heckmi@hhu.de"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Nurul","family":"Lubis","sequence":"additional","affiliation":[{"name":"Heinrich Heine University D\u00fcsseldorf, Germany. lubis@hhu.de"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Carel van","family":"Niekerk","sequence":"additional","affiliation":[{"name":"Heinrich Heine University D\u00fcsseldorf, Germany. niekerk@hhu.de"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Shutong","family":"Feng","sequence":"additional","affiliation":[{"name":"Heinrich Heine University D\u00fcsseldorf, Germany. fengs@hhu.de"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Christian","family":"Geishauser","sequence":"additional","affiliation":[{"name":"Heinrich Heine University D\u00fcsseldorf, Germany. geishaus@hhu.de"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Hsien-Chin","family":"Lin","sequence":"additional","affiliation":[{"name":"Heinrich Heine University D\u00fcsseldorf, Germany. linh@hhu.de"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Milica","family":"Ga\u0161i\u0107","sequence":"additional","affiliation":[{"name":"Heinrich Heine University D\u00fcsseldorf, Germany. gasic@hhu.de"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"281","published-online":{"date-parts":[[2022,11,7]]},"reference":[{"key":"2022110820595851800_bib1","article-title":"Layer normalization","author":"Ba","year":"2016","journal-title":"CoRR"},{"key":"2022110820595851800_bib2","article-title":"Neural machine translation by jointly learning to align and translate","volume-title":"Proceedings of the 3rd International Conference on Learning Representations","author":"Bahdanau","year":"2015"},{"key":"2022110820595851800_bib3","doi-asserted-by":"publisher","first-page":"2476","DOI":"10.21437\/Interspeech.2017-518","article-title":"Towards zero- shot frame semantic parsing for domain scaling","volume-title":"Proceedings of Interspeech 2017","author":"Bapna","year":"2017"},{"key":"2022110820595851800_bib4","doi-asserted-by":"publisher","first-page":"5016","DOI":"10.18653\/v1\/D18-1547","article-title":"MultiWOZ - A large-scale multi-domain Wizard- of-Oz dataset for task-oriented dialogue modelling","volume-title":"Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing","author":"Budzianowski","year":"2018"},{"key":"2022110820595851800_bib5","doi-asserted-by":"publisher","first-page":"122","DOI":"10.18653\/v1\/2020.acl-main.12","article-title":"Zero- shot transfer learning with synthesized data for multi-domain dialogue state tracking","volume-title":"Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics","author":"Campagna","year":"2020"},{"key":"2022110820595851800_bib6","doi-asserted-by":"publisher","first-page":"1468","DOI":"10.21437\/Interspeech.2019-1355","article-title":"BERT-DST: Scalable end-to-end dialogue state tracking with bidirectional encoder representations from transformer","volume-title":"Proceedings of Interspeech 2019","author":"Chao","year":"2019"},{"key":"2022110820595851800_bib7","article-title":"CheckDST: Measuring real-world generalization of dialogue state tracking performance","author":"Cho","year":"2021","journal-title":"CoRR"},{"key":"2022110820595851800_bib8","doi-asserted-by":"publisher","first-page":"127","DOI":"10.1037\/10096-006","article-title":"Grounding in communication","volume-title":"Perspectives on Socially Shared Cognition","author":"Clark","year":"1991"},{"key":"2022110820595851800_bib9","first-page":"879","article-title":"Preview, attend and review: Schema- aware curriculum learning for multi-domain dialogue state tracking","volume-title":"Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 2: Short Papers)","author":"Dai","year":"2021"},{"issue":"1","key":"2022110820595851800_bib10","doi-asserted-by":"publisher","first-page":"755","DOI":"10.1007\/s10462-020-09866-x","article-title":"Survey on evaluation methods for dialogue systems","volume":"54","author":"Deriu","year":"2021","journal-title":"Artificial Intelligence Review"},{"key":"2022110820595851800_bib11","first-page":"4171","article-title":"BERT: Pre-training of deep bidirectional transformers for language understanding","volume-title":"Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long and Short Papers)","author":"Devlin","year":"2019"},{"issue":"8\u20139","key":"2022110820595851800_bib12","doi-asserted-by":"publisher","first-page":"630","DOI":"10.1016\/j.specom.2008.04.002","article-title":"Towards human- like spoken dialogue systems","volume":"50","author":"Edlund","year":"2008","journal-title":"Speech Communication"},{"key":"2022110820595851800_bib13","first-page":"422","article-title":"MultiWOZ 2.1: A consolidated multi-domain dialogue dataset with state corrections and state tracking baselines","volume-title":"Proceedings of the 12th Language Resources and Evaluation Conference","author":"Eric","year":"2020"},{"key":"2022110820595851800_bib14","doi-asserted-by":"crossref","first-page":"264","DOI":"10.18653\/v1\/W19-5932","article-title":"Dialog state tracking: A neural reading comprehension approach","volume-title":"Proceedings of the 20th Annual Meeting of the Special Interest Group on Discourse and Dialogue","author":"Gao","year":"2019"},{"key":"2022110820595851800_bib15","first-page":"781","article-title":"ES-Port: A spontaneous spoken human-human technical support corpus for dialogue research in Spanish","volume-title":"Proceedings of the 11th International Conference on Language Resources and Evaluation","author":"Garc\u00eda-Sardi\u00f1a","year":"2018"},{"key":"2022110820595851800_bib16","doi-asserted-by":"publisher","first-page":"206","DOI":"10.1007\/978-3-030-88483-3_16","article-title":"MultiWOZ 2.3: A multi-domain task-oriented dialogue dataset enhanced with annotation corrections and co-reference annotation","volume-title":"CCF International Conference on Natural Language Processing and Chinese Computing","author":"Han","year":"2021"},{"key":"2022110820595851800_bib17","doi-asserted-by":"publisher","first-page":"6767","DOI":"10.18653\/v1\/2020.coling-main.596","article-title":"Out-of- task training for dialog state tracking models","volume-title":"Proceedings of the 28th International Conference on Computational Linguistics","author":"Heck","year":"2020"},{"key":"2022110820595851800_bib18","doi-asserted-by":"crossref","first-page":"35","DOI":"10.18653\/v1\/2020.sigdial-1.4","article-title":"TripPy: A triple copy strategy for value independent neural dialog state tracking","volume-title":"Proceedings of the 21st Annual Meeting of the Special Interest Group on Discourse and Dialogue","author":"Heck","year":"2020"},{"key":"2022110820595851800_bib19","article-title":"Gaussian error linear units (GELUs)","author":"Hendrycks","year":"2016","journal-title":"CoRR"},{"key":"2022110820595851800_bib20","doi-asserted-by":"publisher","first-page":"17","DOI":"10.3389\/fpsyg.2011.00017","article-title":"Get the story straight: Contextual repetition promotes word learning from storybooks","volume":"2","author":"Horst","year":"2011","journal-title":"Frontiers in Psychology"},{"issue":"4","key":"2022110820595851800_bib21","doi-asserted-by":"publisher","first-page":"1214","DOI":"10.3758\/s13423-015-0980-7","article-title":"The influence of contextual diversity on word learning","volume":"23","author":"Johns","year":"2016","journal-title":"Psychonomic Bulletin & Review"},{"key":"2022110820595851800_bib22","doi-asserted-by":"crossref","first-page":"567","DOI":"10.18653\/v1\/2020.acl-main.53","article-title":"Efficient dialogue state tracking by selectively overwriting memory","volume-title":"Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics","author":"Kim","year":"2020"},{"key":"2022110820595851800_bib23","article-title":"Adam: A method for stochastic optimization","volume-title":"Proceedings of the 3rd International Conference on Learning Representations","author":"Kingma","year":"2015"},{"key":"2022110820595851800_bib24","doi-asserted-by":"publisher","first-page":"8107","DOI":"10.1609\/aaai.v34i05.6322","article-title":"MA-DST: Multi-attention based scalable dialog state tracking","volume-title":"Proceedings of the 34th AAAI Conference on Artificial Intelligence","author":"Kumar","year":"2020"},{"issue":"3\u20134","key":"2022110820595851800_bib25","doi-asserted-by":"publisher","first-page":"323","DOI":"10.1017\/S1351324900002539","article-title":"Information state and dialogue management in the TRINDI dialogue move engine toolkit","volume":"6","author":"Larsson","year":"2000","journal-title":"Natural Language Engineering"},{"key":"2022110820595851800_bib26","doi-asserted-by":"crossref","first-page":"5478","DOI":"10.18653\/v1\/P19-1546","article-title":"SUMBT: Slot-utterance matching for universal and scalable belief tracking","volume-title":"Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics","author":"Lee","year":"2019"},{"key":"2022110820595851800_bib27","article-title":"CoCo: Controllable counterfactuals for evaluating dialogue state trackers","volume-title":"International Conference on Learning Representations","author":"Li","year":"2020"},{"key":"2022110820595851800_bib28","first-page":"1063","article-title":"Zero-shot generalization in dialog state tracking through generative question answering","volume-title":"Proceedings of the 16th Conference of the European Chapter of the Association for Computational Linguistics: Main Volume","author":"Li","year":"2021"},{"key":"2022110820595851800_bib29","article-title":"Attention guided dialogue state tracking with sparse supervision","author":"Liang","year":"2021","journal-title":"CoRR"},{"key":"2022110820595851800_bib30","doi-asserted-by":"crossref","first-page":"7890","DOI":"10.18653\/v1\/2021.emnlp-main.622","article-title":"Zero-shot dialogue state tracking via cross- task transfer","volume-title":"Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing","author":"Lin","year":"2021"},{"key":"2022110820595851800_bib31","doi-asserted-by":"publisher","first-page":"2506","DOI":"10.21437\/Interspeech.2017-1326","article-title":"An end-to-end trainable neural network model with belief tracking for task-oriented dialog","volume-title":"Proceedings of Interspeech 2017","author":"Liu","year":"2017"},{"key":"2022110820595851800_bib32","article-title":"RoBERTa: A robustly optimized BERT pretraining approach","volume":"abs\/1907.11692v1","author":"Liu","year":"2019","journal-title":"CoRR"},{"key":"2022110820595851800_bib33","article-title":"An end-to-end dialogue state tracking system with machine reading comprehension and wide & deep classification","volume":"abs\/1912.09297v2","author":"Ma","year":"2019","journal-title":"CoRR"},{"key":"2022110820595851800_bib34","article-title":"DialoGLUE: A natural language understanding benchmark for task-oriented dialogue","author":"Mehri","year":"2020","journal-title":"CoRR"},{"key":"2022110820595851800_bib35","doi-asserted-by":"publisher","first-page":"1777","DOI":"10.18653\/v1\/P17-1163","article-title":"Neural belief tracker: Data-driven dialogue state tracking","volume-title":"Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)","author":"Mrk\u0161i\u0107","year":"2017"},{"key":"2022110820595851800_bib36","doi-asserted-by":"publisher","first-page":"7803","DOI":"10.1109\/ICASSP39728.2021.9413810","article-title":"Language model is all you need: Natural language understanding as question answering","volume-title":"Proceedings of the 2021 IEEE International Conference on Acoustics, Speech and Signal Processing","author":"Namazifar","year":"2021"},{"key":"2022110820595851800_bib37","article-title":"Recent advances in deep learning based dialogue systems: A systematic survey","volume":"abs\/2105.04387v4","author":"Ni","year":"2021","journal-title":"CoRR"},{"key":"2022110820595851800_bib38","article-title":"Toward scalable neural dialogue state tracking model","volume-title":"Proceedings of the 2nd Conversational AI workshop at the 32nd Conference on Neural Information Processing Systems","author":"Nouri","year":"2018"},{"key":"2022110820595851800_bib39","doi-asserted-by":"crossref","first-page":"326","DOI":"10.18653\/v1\/2021.sigdial-1.35","article-title":"Annotation inconsistency and entity bias in MultiWOZ","volume-title":"Proceedings of the 22nd Annual Meeting of the Special Interest Group on Discourse and Dialogue","author":"Qian","year":"2021"},{"key":"2022110820595851800_bib40","first-page":"82","article-title":"Text chunking using transformation-based learning","volume-title":"Proceedings of the 3rd Workshop on Very Large Corpora","author":"Ramshaw","year":"1995"},{"key":"2022110820595851800_bib41","article-title":"Schema-guided dialogue state tracking task at DSTC8","author":"Rastogi","year":"2020","journal-title":"CoRR"},{"key":"2022110820595851800_bib42","doi-asserted-by":"publisher","first-page":"8689","DOI":"10.1609\/aaai.v34i05.6394","article-title":"Towards scalable multi-domain conversational agents: The schema-guided dialogue dataset","volume-title":"Proceedings of the 34th AAAI Conference on Artificial Intelligence","author":"Rastogi","year":"2020"},{"key":"2022110820595851800_bib43","doi-asserted-by":"crossref","first-page":"2780","DOI":"10.18653\/v1\/D18-1299","article-title":"Towards universal dialogue state tracking","volume-title":"Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing","author":"Ren","year":"2018"},{"key":"2022110820595851800_bib44","doi-asserted-by":"publisher","first-page":"44","DOI":"10.18653\/v1\/W18-5606","article-title":"Deep learning for language understanding of mental health concepts derived from cognitive behavioural therapy","volume-title":"Proceedings of the 9th International Workshop on Health Text Mining and Information Analysis","author":"Rojas-Barahona","year":"2018"},{"key":"2022110820595851800_bib45","volume-title":"Image Analysis and Mathematical Morphology","author":"Serra","year":"1982"},{"key":"2022110820595851800_bib46","article-title":"Building a conversational agent overnight with dialogue self-play","author":"Shah","year":"2018","journal-title":"CoRR"},{"key":"2022110820595851800_bib47","first-page":"4661","article-title":"Multi-task pre-training for plug-and- play task-oriented dialogue system","volume-title":"Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)","author":"Yixuan","year":"2022"},{"key":"2022110820595851800_bib48","doi-asserted-by":"publisher","first-page":"80","DOI":"10.18653\/v1\/2021.nlp4convai-1.8","article-title":"Amendable generation for dialogue state tracking","volume-title":"Proceedings of the 3rd Workshop on Natural Language Processing for Conversational AI","author":"Tian","year":"2021"},{"key":"2022110820595851800_bib49","first-page":"5998","article-title":"Attention is all you need","volume-title":"Advances in Neural Information Processing Systems","author":"Vaswani","year":"2017"},{"key":"2022110820595851800_bib50","first-page":"438","article-title":"A network-based end-to-end trainable task-oriented dialogue system","volume-title":"Proceedings of the 15th Conference of the European Chapter of the Association for Computational Linguistics: Volume 1, Long Papers","author":"Wen","year":"2017"},{"key":"2022110820595851800_bib51","first-page":"917","article-title":"TOD-BERT: Pre-trained natural language understanding for task-oriented dialogue","volume-title":"Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing","author":"Chien-Sheng","year":"2020"},{"key":"2022110820595851800_bib52","first-page":"808","article-title":"Transferable multi- domain state generator for task-oriented dialogue systems","volume-title":"Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics","author":"Chien-Sheng","year":"2019"},{"key":"2022110820595851800_bib53","article-title":"Google\u2019s neural machine translation system: Bridging the gap between human and machine translation","author":"Yonghui","year":"2016","journal-title":"CoRR"},{"key":"2022110820595851800_bib54","first-page":"1448","article-title":"An end-to-end approach for handling unknown slot values in dialogue state tracking","volume-title":"Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)","author":"Puyang","year":"2018"},{"key":"2022110820595851800_bib55","first-page":"258","article-title":"Targeted feature dropout for robust slot filling in natural language understanding","volume-title":"Fifteenth Annual Conference of the International Speech Communication Association","author":"Puyang","year":"2014"},{"key":"2022110820595851800_bib56","article-title":"MultiWOZ 2.4: A multi- domain task-oriented dialogue dataset with essential annotation corrections to improve state tracking evaluation","author":"Ye","year":"2021","journal-title":"CoRR"},{"key":"2022110820595851800_bib57","doi-asserted-by":"crossref","first-page":"1598","DOI":"10.1145\/3442381.3449939","article-title":"Slot self-attentive dialogue state tracking","volume-title":"Proceedings of the Web Conference 2021","author":"Ye","year":"2021"},{"issue":"2","key":"2022110820595851800_bib58","doi-asserted-by":"publisher","first-page":"150","DOI":"10.1016\/j.csl.2009.04.001","article-title":"The hidden information state model: A practical framework for POMDP-based spoken dialogue management","volume":"24","author":"Young","year":"2010","journal-title":"Computer Speech & Language"},{"key":"2022110820595851800_bib59","article-title":"SCoRe: Pre-training for context representation in conversational semantic parsing","volume-title":"International Conference on Learning Representations","author":"Tao","year":"2020"},{"key":"2022110820595851800_bib60","doi-asserted-by":"publisher","first-page":"109","DOI":"10.18653\/v1\/2020.nlp4convai-1.13","article-title":"MultiWOZ 2.2: A dialogue dataset with additional annotation corrections and state tracking baselines","volume-title":"Proceedings of the 2nd Workshop on Natural Language Processing for Conversational AI","author":"Zang","year":"2020"},{"key":"2022110820595851800_bib61","first-page":"154","article-title":"Find or classify? Dual strategy for slot-value predictions on multi-domain dialog state tracking","volume-title":"Proceedings of the 9th Joint Conference on Lexical and Computational Semantics","author":"Zhang","year":"2020"},{"key":"2022110820595851800_bib62","doi-asserted-by":"publisher","first-page":"2011","DOI":"10.1007\/s11431-020-1692-3","article-title":"Recent advances and challenges in task- oriented dialog systems","volume":"63","author":"Zhang","year":"2020","journal-title":"Science China Technological Sciences"},{"key":"2022110820595851800_bib63","first-page":"3740","article-title":"Modeling multi-turn conversation with deep utterance aggregation","volume-title":"Proceedings of the 27th International Conference on Computational Linguistics","author":"Zhang","year":"2018"},{"key":"2022110820595851800_bib64","doi-asserted-by":"publisher","first-page":"1458","DOI":"10.18653\/v1\/P18-1135","article-title":"Global-locally self-attentive encoder for dialogue state tracking","volume-title":"Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)","author":"Zhong","year":"2018"}],"container-title":["Transactions of the Association for Computational Linguistics"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/direct.mit.edu\/tacl\/article-pdf\/doi\/10.1162\/tacl_a_00513\/2057218\/tacl_a_00513.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"syndication"},{"URL":"https:\/\/direct.mit.edu\/tacl\/article-pdf\/doi\/10.1162\/tacl_a_00513\/2057218\/tacl_a_00513.pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2023,11,30]],"date-time":"2023-11-30T12:59:43Z","timestamp":1701349183000},"score":1,"resource":{"primary":{"URL":"https:\/\/direct.mit.edu\/tacl\/article\/doi\/10.1162\/tacl_a_00513\/113662\/Robust-Dialogue-State-Tracking-with-Weak"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2022]]},"references-count":64,"URL":"https:\/\/doi.org\/10.1162\/tacl_a_00513","relation":{},"ISSN":["2307-387X"],"issn-type":[{"value":"2307-387X","type":"electronic"}],"subject":[],"published-other":{"date-parts":[[2022]]},"published":{"date-parts":[[2022]]}}}